What is the elegant way of doing resampling with replacement in a group_by/summarise workflow?

[Leon]

I want to do repeated resampling with replacement, which can be done like the generic example below, but I'm wondering if there is a more elegant solution than the one applied in the lapply()-step?

# Load libraries ----------------------------------------------------------
library("tidyverse")


# Define example data -----------------------------------------------------
n <- 1000
my_data <- tibble(
  x = rnorm(n),
  y = rnorm(n),
  group = sample(LETTERS[1:4],
                 size = n,
                 replace = TRUE),
  class = sample(c(0, 1),
                 size = n,
                 replace = TRUE)
)


# Resample data once ------------------------------------------------------
set.seed(859948)
my_data %>% 
  group_by(group, class) %>% 
  sample_n(100,
           replace = TRUE) %>% 
  ungroup %>% 
  group_by(group, class) %>%
  summarise(value = mean(x) - mean(y))


# Resample data multiple times --------------------------------------------
set.seed(859948)
my_resamples <- lapply(1:1000,
                       function(i){
                         my_data %>% 
                           group_by(group, class) %>% 
                           sample_n(100,
                                    replace = TRUE) %>% 
                           ungroup %>% 
                           group_by(group, class) %>%
                           summarise(value = mean(x) - mean(y)) })
my_resamples <- my_resamples %>% 
  bind_rows


# Visualise distributions -------------------------------------------------
my_resamples %>%
  ggplot(aes(x = value,
             y = group,
             fill = factor(class))) +
  geom_boxplot() +
  theme_minimal()

Yielding